Patients with clinically significant heart failure comprise the largest and most rapidly growing subset of patients with cardiovascular disease in the U.S. The burden in regard to morbidity, mortality, and health care expense is staggering. In the clinical management of critically ill new-onset heart failure patients, the inability to immediately differentiate those who will respond to medical therapy from those who will not results in the inappropriate application of high-risk heart failure surgery in patients who do not need it, while also denying surgery to those who have the most to benefit from it. MRI-based multi-parametric strain analysis generates a patient-specific, normalized, and highly accurate description of micro-regional left ventricular (LV) contractile function by comparing three patient specific raw strain parameter values from each of 15,300 grid points (a total of 45,900 database comparisons per patient) to the respective normal average +/- standard deviation (from our laboratory's normal human strain database; n>70). This normalization of multiple different patient- specific strain parameter values allows their combination into novel and powerful multi-parametric strain indices that have proven more powerful than any individual strain metric. By applying MRI-based multi-parametric strain analysis in new-onset heart failure (dilated cardiomyopathy) patients, we have discovered marked heterogeneity in the LV regional distribution of contractile injury. This unexpected heterogeneity of contractile injury in dilated cardiomyopathy patients occurs in a consistent pattern that has allowed us to identify sentinel LV regions (basilar-septal and mid-septal) that are more sensitive to early injury and that serve as harbingers of irreversible contractile injury in new-onset heart failure patients. We hypothesize that the surveillance of these sentinel LV regions by MRI-based multi-parametric strain analysis allows the risk stratification of new onset heart failure patients in regard to eary response to medical therapy and the subsequent occurrence of heart failure adverse events. New-onset heart failure patients who 1) do not yet have a cardioverter/defibrillator or biventricular pacemaker, 2) have an ejection fraction <30%, and 3) have idiopathic dilated cardiomyopathy (no coronary artery or valvular disease) will undergo MRI-based multi- parametric strain analysis to produce average Z-scores for 72 LV micro-regions. After stabilization on heart failure medical therapy (ACC/AHA Clinical Guidelines), they will be restudied 3 months later. We will statistically characterize the ability of baseline basilar-septal and mid-septal sentinel regional average multi- parametric strain Z-scores to predict both contractile recovery on medical therapy and the occurrence of heart failure adverse events. We will statistically optimize the micro-regional makeup of our sentinel surveillance region and compare our results to the predictive capabilities of the current standard, echocardiographic ejection fraction. A metric of micro-regional LV contractile function that predicts early recovery on heart failure medical therapy would more accurately direct the application of high-risk surgical therapy for heart failure.